Skip to main content

Markdown task board for your repo

Project description

.mdboard

A lightweight kanban board where every task is a .md file in your repo.

Built for solo developers and AI-assisted workflows. No database, no SaaS, no dependencies -- just markdown files that live alongside your code.

Why not a cloud board?

Cloud project management tools are built for teams coordinating across time zones on work that stretches over days and weeks. But when it's you and your AI agents on one machine, shipping tasks that take minutes to hours, a cloud subscription is overhead. You're paying for collaboration features you don't use, managing another account per project, and context-switching to a browser tab that's disconnected from the code you're actually working on.

mdboard puts the board in the repo. Tasks are markdown files checked into git. No subscription, no sync issues, no per-project setup. Clone the repo and the board is already there.

The board travels with the branch. Your main branch has a roadmap. Your experiment/new-parser branch has its own set of tasks that don't exist on main. When you merge, the task history merges too. When you abandon the branch, the tasks disappear with it. The tasks/ directory becomes a built-in audit trail of what was planned, what was done, and what was left behind -- scoped to the exact line of work.

Markdown is the universal interface. Humans read it. LLMs read it. Git diffs it. Your editor previews it. No proprietary format, no API to learn -- just files.

AI agents need a local work surface. Give an agent a tasks/ directory and it can create work items, track progress, check off acceptance criteria, and leave an audit trail -- all through basic file operations. No auth tokens, no external services, no context window wasted on API docs.

Quick start

uvx mdboard

Open http://localhost:8080.

How it works

Tasks are markdown files with YAML frontmatter. Columns are directories. Moving a task = moving a file.

tasks/
├── backlog/
├── todo/
│   └── 002-setup-ci-pipeline.md
├── in-progress/
│   └── 004-add-api-rate-limiting.md
├── review/
├── done/
│   └── 001-implement-user-authentication.md
└── comments/

A task file:

---
id: 2
title: Setup CI pipeline
assignee: claude
tags: [devops, ci]
created: 2026-02-08
branch: feature/ci
---

## Description
Configure GitHub Actions for linting, testing, and deployment.

## Acceptance Criteria
- [ ] Lint on every push
- [x] Run test suite on every PR
- [ ] Auto-deploy to staging on merge to main

## Notes
Decided to use composite actions for reusability.

You can manage tasks from the web UI, from the command line, or from an AI agent -- they're all just reading and writing files.

# move a task to in-progress
mv tasks/todo/002-setup-ci-pipeline.md tasks/in-progress/

# create a task from the command line
cat > tasks/backlog/007-fix-login-bug.md << 'EOF'
---
id: 7
title: Fix login bug
assignee: claude
tags: [bug]
created: 2026-02-08
---

## Description
Login fails when password contains special characters.
EOF

AI agent workflow

The board is designed to be operated by AI coding agents. A typical setup:

  1. Agent scans tasks/todo/ for tasks assigned to it
  2. Moves the task file to tasks/in-progress/
  3. Does the work, checking off acceptance criteria as it goes
  4. Appends notes about decisions and changes made
  5. Moves the task to tasks/done/ and adds completed: 2026-02-08 to frontmatter
  6. Creates new task files in tasks/backlog/ for anything discovered along the way

Everything is committed alongside the code changes. The git log tells you what was done and why.

Features

  • Zero dependencies -- Python standard library only
  • Git-native -- tasks are version-controlled with your code
  • Branch-aware -- each branch carries its own task state
  • Drag-and-drop UI -- web-based kanban board
  • Filtering -- by assignee or tag
  • Inline editing -- edit raw markdown in the browser
  • Checkbox tracking -- card progress from - [x] items
  • Comments -- per-task comment threads, also stored as markdown
  • Keyboard shortcuts -- N new task, Esc close

Server options

mdboard --port 3000           # custom port
mdboard --tasks-dir ./other   # different tasks directory

API

JSON API for scripting:

Method Endpoint Description
GET /api/board Full board state
POST /api/task Create a task
PUT /api/task/{col}/{file} Update a task
DELETE /api/task/{col}/{file} Delete a task
PATCH /api/task/move Move between columns
GET/POST/DELETE /api/comments/{id} Task comments

Requirements

Python 3.12+

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

mdboard-0.1.0.tar.gz (21.2 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

mdboard-0.1.0-py3-none-any.whl (19.7 kB view details)

Uploaded Python 3

File details

Details for the file mdboard-0.1.0.tar.gz.

File metadata

  • Download URL: mdboard-0.1.0.tar.gz
  • Upload date:
  • Size: 21.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.9.28 {"installer":{"name":"uv","version":"0.9.28","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for mdboard-0.1.0.tar.gz
Algorithm Hash digest
SHA256 4f35da37b7007405eb8c35097129dde5481bb9f138ef57b3bc56ea7897bec598
MD5 a2c83444378c811e2d85b1c8d1b07af6
BLAKE2b-256 34d8ee38de01b520d92fa594c20d51e93ef2bf8e1f85fe01b47f3783cb35410a

See more details on using hashes here.

File details

Details for the file mdboard-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: mdboard-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 19.7 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.9.28 {"installer":{"name":"uv","version":"0.9.28","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"macOS","version":null,"id":null,"libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}

File hashes

Hashes for mdboard-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 95b75a3e9f75b7f71c853116430ce4316d1befd326f91b30796857f0abc19319
MD5 a5b1a068e571f9db5c6a876aeacf7fe9
BLAKE2b-256 e21b7cef33961b3aa4990d5a09d7f5098cb9716c4f4485ffee14039cd60115f7

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page